How to Integrate IBM Watson with Intercom

Set Up IBM Cloud Account and Services

• Create an IBM Cloud account. If you already have one, log in to your dashboard.

 

• Navigate to the IBM Cloud Catalog and find Watson Assistant by searching for it in the AI services section.

 

• Provision a Watson Assistant service instance. Follow the prompts to create and name your instance.

 

• Access the Watson Assistant service dashboard once your instance is ready. Take note of your API Key and Assistant ID as you will need them later.

Configure Intercom Account

• If you do not have an Intercom account, sign up for one and log into the dashboard.

 

• Set up a new App on Intercom if you don't already have one. Go to the "App settings" section to retrieve your Intercom API Key.

Develop the Integration using Node.js

• Ensure Node.js and npm are installed on your system. If not, download and install them from the official Node.js website.

 

• Create a new Node.js project:

 

mkdir ibm-watson-intercom-integration
cd ibm-watson-intercom-integration
npm init -y

• Install necessary npm packages for IBM Watson SDK and Intercom client:

npm install ibm-watson intercom-client express body-parser

Write the Server Code

• Create an `index.js` file and set up a basic Express server:

const express = require('express');
const bodyParser = require('body-parser');
const AssistantV2 = require('ibm-watson/assistant/v2');
const Intercom = require('intercom-client');

const app = express();
app.use(bodyParser.json());
const port = 3000;

app.listen(port, () => {
  console.log(`Server running on port ${port}`);
});

• Initialize Watson Assistant and Intercom clients using your credentials:

const assistant = new AssistantV2({
  version: '2021-06-14',
  authenticator: new iamAuthenticator({
    apikey: 'your-watson-api-key',
  }),
  serviceUrl: 'https://api.us-south.assistant.watson.cloud.ibm.com',
});

const intercomClient = new Intercom.Client({ token: 'your-intercom-api-key' });

• Add an endpoint to handle incoming messages from Intercom and forward them to Watson Assistant:

app.post('/webhook', async (req, res) => {
  const userMessage = req.body.message;
  
  try {
    const session = await assistant.createSession({ assistantId: 'your-assistant-id' });

    const response = await assistant.message({
      assistantId: 'your-assistant-id',
      sessionId: session.result.session_id,
      input: {
        'message_type': 'text',
        'text': userMessage,
      },
    });

    const assistantReply = response.result.output.generic[0].text;

    await intercomClient.messages.create({
      from: { type: 'user', id: req.body.user_id },
      body: assistantReply,
    });

    res.sendStatus(200);
  } catch (err) {
    console.error(err);
    res.sendStatus(500);
  }
});

Deploy and Test the Integration

• Host your Node.js application. For local testing, you can use tools like ngrok to expose your localhost.

 

• Set up a webhook in your Intercom settings to point to your server's `/webhook` endpoint URL.

 

• Communicate via your Intercom app and observe responses from Watson Assistant to ensure the integration works as expected.

Refine and Secure Your Integration

• Ensure secure data transmission between your services by using HTTPS and managing API key confidentiality.

 

• Refactor and modularize code for better maintainability and adhere to best practices for product readiness.